popcorn in the cinema 1 running head: popcorn and ... · half received 65 gram popcorn or a chewing...

POPCORN IN THE CINEMA 1

Running head: Popcorn and advertising

Popcorn in the Cinema: Oral Interference Sabotages Advertising Effects

Sascha Topolinski, Sandy Lindner, & Anna-Lena Freudenberg

University of Wuerzburg

IN PRESS, JOURNAL OF CONSUMER PSYCHOLOGY

Word count: 5,610 (main text) Please address correspondence to:

Sascha Topolinski

Department of Psychology, Social and Economic Cognition

University of Cologne, Richard-Strauß-Straße 2, 50931, Cologne

E-mail: [email protected]


Abstract

One important psychological mechanism of advertising is mere exposure inducing positive

attitudes towards brands. Recent basic research has shown that the underlying mechanism of

mere exposure for words, in turn, is the training of subvocal pronunciation, which can be

obstructed by oral motor-interference. Commercials for foreign brands were shown in cinema

sessions while participants either ate popcorn, chewed gum (oral interference) or consumed a

single sugar cube (control). Brand choice and brand attitudes were assessed one week later.

While control participants more likely spent money (Experiment 1, N = 188) and exhibited

higher preference and physiological responses (Experiment 2, N = 96) for advertised than for

novel brands, participants who had consumed popcorn or gum during commercials showed no

advertising effects. It is concluded that advertising might be futile under ecological situations

involving oral interference, such as snacking or talking, which ironically is often the case.

145 words

Key words: advertising, embodiment, persuasion, marketing, mere exposure


Popcorn in the Cinema: Oral Interference Sabotages Advertising Effects

The main purpose of marketing and advertising is to induce a more favorable brand

attitude thereby increasing the likelihood of eventual brand choice (e.g., Baker, 1999; Elliott,

& Yannopoulou, 2007). A central psychological mechanism of the impact of advertising on

brand choice is the mere exposure effect (for a review, see Grimes & Kitchen, 2007). This

effect, well-established in experimental psychology (Bornstein, 1989), is the phenomenon that

any sort of stimulus is preferred when it is repeatedly presented (Zajonc, 1968). This

preference in turn is due to higher processing fluency, that is, increased efficiency of

information processing, of repeated compared to novel stimuli (Reber, Winkielman, &

Schwarz, 1998), with earlier research having shown that high fluency per se feels generally

positive (Reber, Schwarz, & Winkielman, 2004; Topolinski & Reber, 2010; Topolinski &

Strack, 2009a, 2009c; Topolinski, Likowski, Weyers, & Strack, 2009). In that sense,

advertising is simply a method to repeat brands and thus increase the easiness and joy in

mentally processing a brand name. Indeed, earlier research has shown that mere exposure of

brands actually increases positive attitudes and the likelihood of eventual brand choice (e.g.,

Baker, 1999; Blüher & Pahl, 2007; Janiszewski, 1993; Matthes, Schemer, & Wirth, 2007;

Lodish, Magid, Kalmenson, Livelsberger, & Lubetkin, 1995).

Recent basic research has shown that the causal mechanism of the mere exposure

effect, in turn, is the fluency of covert stimulus-specific motor-simulations (Topolinski &

Strack, 2009b). The underlying rational for this was the following. For words, for instance,

each time a word is encountered, a covert simulation of pronouncing the word takes place (cf.,

Stroop, 1935). When the word is encountered repeatedly, this covert pronouncing simulation

is also repeated and thus runs more fluently for repeated compared to novel words. This gain

in oral motor-fluency triggers a positive feeling that drives mere exposure effects (see also

Leder, Bär, & Topolinski, 2013; Moreland & Topolinski, 2011; Topolinski, 2010). However,

when the oral motor-system is prevented from training such sub-vocalizations, for instance by


merely chewing gum, there is no gain in pronunciation simulation fluency for old over novel

words and thus no exposure effects for words (Topolinski & Strack, 2009b, 2010).

Testing this oral embodiment account of the mere exposure effect, Topolinski and

Strack (2009b) presented nonsense words and Chinese ideographs (as a control condition) to

participants –with a random half of these stimuli repeated– and asked how much the

participants liked each of these stimuli. Crucially, two kinds of motor interference were

implemented. A control group simply kneaded a ball as a manual interference. This should

introduce some distraction but left the mouth free to covertly simulate the pronunciation of

the names. A crucial experimental group, however, chewed gum during the presentation of

the stimuli, which should prevent their mouth from simulating the words’ pronunciation. The

result was that the manual group preferred repeated over novel names, but the oral motor-

interference group did not. In contrast, both groups preferred repeated over novel ideographs,

obviously because neither the manual nor the oral secondary task interfered with the merely

visual encoding of these images.

Moreover, Topolinski and Strack (2009b, Experiment 3) investigated a double

dissociation between two sorts of stimuli and two sorts of simulation modality. Specifically,

in another experiment they presented (partially repeated) words and tunes. While words are to

be spoken, tunes are to be sung. Thus, it was hypothesized that a mere exposure of a tune

leads to a singing or humming simulation in the vocal folds. Testing this, humming as a vocal

interference was implemented that should prevent the vocal folds from covertly simulating the

voice pitch variations of the tunes but leave the mouth free to simulate word pronunciations.

In contrast, a tongue movement exercise was implemented as a purely oral interference that

should (such as chewing gum) prevent the mouth from simulating word pronunciations but

leave the vocal folds free to simulate the tunes. And that was actually what was found:

participants in the vocal interference condition showed a mere exposure effect for words, but


not for tunes, while participants in the oral interference condition showed a mere exposure

effect for tunes, but not for words.

In sum, these recent findings support the notion that mere exposure effects hinge on

the fluency of stimulus-related motor-simulations. Furthermore, since repetition is only one of

many ways to increase processing fluency (Reber, Schwarz, & Winkielman, 2004), the same

should be expected for other fluency manipulations. For instance, Song & Schwarz (2009)

presented relatively easy (e.g., Magnalroxate) or hard (e.g., Hnegripitrom) to pronounce

names of ostensible food additives. In this case, the fluency does not stem from repetition but

directly from pronunciation itself. They found that participants rated the easy-to-pronounce

additives as being less harmfull than hard-to-pronounce ones. Obviously, participants had

based their judgments on the easiness of pronouncing a given name. Testing whether also this

fluency is orally embodied, Topolinski and Strack (2010, Experiment 3) replicated the

experiment under manual and oral interference and found the pronunciation-easiness effect

for manual interference only.

Although the multidimensionality of sources determining brand attitudes has long

been researched (e.g., Keller, 2003) and there have been various recent advances into the

underlying cognitive and affective mechanisms in brand attitude formation (e.g., Esch, Möll,

Schmitt, Elger, Neuhaus, & Weber, 2012; Reimann, Castaño, Zaichkowsky & Bechara, 2012;

Schmitt, 2012; Venkatraman, Clithero, Fitzsimons, & Huettel, 2012) this embodied source of

oral motor-fluency has not been considered to date.

Ironically, many everday life situations involving advertising also involve continuous

oral interference, such as nibbling snacks while watching TV, or eating popcorn in the cinema

while watching the commercials before the main movie. Thus, it is likely that these settings

also sabotage oral fluency gains from mere exposure. Because fluency effects play an

important role in consumer choices (Alter & Oppenheimer, 2006; Novemsky, Dhar, Schwarz,

& Simonson, 2007) we hypothesized that oral interference should also hamper the impact of


exposure on brand attitudes and choice. The present experiments were designed to test this

idea in controlled field studies.

Experiment 1

A cinema session involving eating candy during the presentation of commercials for

competitive sets of unknown brands (body lotions, charity foundations) was held in a

lecturing hall. The likelihood of spending money on the advertised brands (purchasing body

lotions, donating for charity foundations) in an interbrand choice one week later was the

dependent measure. Data collection was divided into two campaigns, each with different

samples, one featuring popcorn and one featuring gum as oral interference. Each of these

campaigns realized as between-subjects design an oral interference (popcorn or gum)

condition and a control condition (always a sugar cube) and consisted of a study phase (the

actual cinema session) and a test phase one week later. Gum was chosen as an additional

instantiation of oral interference to rule out distraction: consuming popcorn does not only

entail oral movements, but also taking the popcorn out of the bag and looking into the bag.

Thus, if chewing gum would also block advertising effects, then the impact of eating popcorn

could not be attributed to distraction. We predicted that participants in the control group

would more likely choose advertised than novel options, while participants with oral motor-

interference would not show such an advertising effect.

Method

Participants. N = 197 female psychology freshmen from a university in their first

week of courses participated for course credit (mean age 21, SD = 3). Nine of them did not

attend the test phase (see below), resulting in N = 188. Since the base rate of women is

generally much higher in psychology courses, only female students were invited. Experiment

2 also involved male participants.


Materials. We used 6 real commercials downloaded from the internet for existing

body lotion brands, namely Arko Nem©, Aveeno©, Emeron©, Puhas Loodus©, Boroplus©,

Innisfree©. The language was always non-German. Pilot studies had shown that these

products were unfamiliar to German participants. Product names were verbally mentioned

within all the commercials. The commercials ranged in length between 15 – 61 seconds (M =

31, SD = 16). Furthermore, we used 6 fictitious names of charity foundations (Sitais Geteref,

Désecona Seltes, Aevenge Etarnes, Aexilieste Jomé, Cirates Cencare, Flendur Kagintes) that

were featured in advertising slides with some ornaments and one catch-phrase (e.g., For a life

against cancer) of 7 seconds presentation length. Filler commercials were for mineral water

brands, lemonades, and potato chips. The short films in test phase were in the popcorn

campaign, Verschollen über den Wolken (21 minutes in length), and Zero (12 minutes in

length), in the gum campaign.

Procedure. In the study phase, groups of 20-25 individuals were brought into a dimly

lit lecturing hall. The audience was divided into a right and a left half (with a viewing screen,

approx. 3 X 5 meters, for each) using viewing partitions that also prevented each group from

seeing what the other group received. Participants were randomly assigned to the halves. One

half received 65 gram popcorn or a chewing gum (oral interference conditions) and the other

half received a sugar cube (control, comparable in hedonic experience and some caloric

input). The assignment of conditions to right and left halves was counter-balanced across

sessions; snacks were delivered in hygienic paper bags to the seats. Due to the viewing

partitions between the halfes and the dim light, participants could not see what the other half

was consuming.

Participants were informed that a typical cinema session would be simulated and that

they would later answer some questions concerning the movie. Participants were instructed to

start consuming their respective snack while three filler commercials (115 seconds in length)

were presented. Compliance with these instructions was observed and enforced by research


assistants. The time of the filler commercials served as a buffer for the control participants to

consume the sole sugar cube. Then the crucial target commercials were shown in a fixed

random order (one random half of the body lotions and of the charity foundations,

respectively; materials counter-balanced across participants) intermixed with further filler

commercials. By the time the target commercials were presented, the control participants had

already consumed their sole sugar cube (as tested in pilots and carefully observed by research

assistants during the session) so that no further oral interference took place for them. The

popcorn group, conversely, took the whole duration of the commercial presentation or more to

consume their popcorn. After the commercials (total length 8 minutes) a short film was shown

and then participants answered some filler items about the movie as well as on mood and

eating behavior on a paper-pencil questionnaire.

The test phase was implemented one week after the study session. Participants

attended the test phase in the same groups as in the earlier cinema study sessions. They were

brought into a café and were given a small paper purse containing 4 €1-coins. They were told

that this was an additional reward for the participation, which they had to spend during this

session. To justify the cover story of product testing and to induce a relaxed consuming mind-

set, they were asked to spend 2 of the 1€ coins for snacks and beverages on the menu (each

product € 1) and freely chat with each other. The menu cards featured some of the advertised

beverages and chips from the filler commercials of the study session to familiarize

participants with the idea that some products from the study session would re-occur and to

reduce their suspiciousness on the later re-occurring lotion and charity foundation names.

After they consumed their orders, participants were brought to two stands involving the

interbrand choices one at a time. The other participants could not see these choices.

Body lotion purchasing. On a stand, 6 plastic bottles filled with 50 ml body lotion

and prepared with stickers depicting the 6 product names (alphabetical order from left to

right), with slightly varying ornaments and colorings were presented together with a money


box for each product. Crucially, half of the names have been advertised in the cinema session

(stimulus assignment counter-balanced across participants). Participants were asked to choose

one bottle and to put €1 into the respective money box, observed by a research assistant.

Donation. On a second stand, 6 donation cans featuring the 6 names of the charity

foundations (alphabetical sequence from left to right) were presented and participants were

asked to read all names and then to drop €1 into one of the boxes. Again, half of these names

had been advertised.

Data assessment. Unbeknownst to the participants, depending on their oral-

interference condition they had received either 1€-coins with a German image on the national

backside of the coin, or a non-German image from another issuing country (assignment of

nationality to interference condition counter-balanced across groups). Due to the high degree

of intermingling of coins from different issuing countries in Europe, this manipulation was

unobtrusive. After the experiment, the number of German and non-German coins in the

respective money boxes were counted. Since assignment of product names to old vs. new

status was the same for one experimental group, and groups of individuals were the same as in

the study sessions, the number of coins assigned to old vs. new products could easily be

assessed by simply counting German and non-German coins in the money boxes.

Results

Because the present data are categorical and coins could not be assigned to

individuals, we ran both parametric and non-parametric analyses, which yielded similar

results. Because they are more common to readers, we first report the more illustrative

parametric analyses.

Parametric analyses. A 2 (oral interference: yes, no; between) X 2 (type of oral

interference: popcorn, gum; between) X 2 (product: buying a lotion, donating for an

organization; within) mixed ANOVA on the likelihood of choosing an advertised option


found a significant main effect of oral interference, F(1, 184) = 26.85, p < .0001, ηp2 = .13.

Thus, we collapsed over product and type of oral interference (gum or popcorn). The group

with no oral interference showed an advertising effect. They chose the advertised products

with a likelihood of 59 % (SD = 0.34, SE = 0.04), which is reliably above the chance level of

50 %, t(91) = 2.42, p = .017, d = 0.26. However, the groups with oral interference showed no

advertising effect. They chose the advertised products with a likelihood of 37 % (SD = 0.22,

SE = 0.04), which was even below chance level, t(96) = 5.78, p < .001, d = 0.59.

As a side result, there was also an interaction between oral interference and type of

oral interference, F(1, 184) = 3.99, p = .047, ηp2 = .02. However, as Figure 1 shows, this

interaction was driven by the conceptually irrelevant fact that the likelihood of chosing an old

lotion was lower in the control group for the popcorn subsample than in the control group for

the gum sub-sample.

Non-parametric analyses. Non-parametric analyses using Chi-square yielded the

same pattern. Participants with no oral interference chose the advertised product more often

than participants in the oral interference group, χ2(1, N = 188) = 7.65, p = .0006. The likehood

of chosing an advertised option was higher than chance level in the group with no oral

interference, χ2(1, N = 92) = 4.38, p = .037, but lower than chance level in the group with oral

interference, χ2(1, N = 96) = 4.17, p = .041.

Discussion

As predicted, while participants in a control condition showed advertising effects, oral

interference due to eating popcorn or chewing gum while watching commercials reduced this

advertising effects for an interbrand choice of initially unfamiliar brands as measured one

week later. This blockade was effective both for purchasing a product and donating money.

Although not predicted, we even found a reduction of consumer choice likelihood due to oral

interference. Such a reversal of preference due to repetition has not yet been observed in


previous studies on oral interference (Topolinski, 2012; Topolinski & Strack, 2009b, 2010)

and was not replicated in Experiment 2. Thus, tentatively we speculate the following possible

mechanism for this. It is possible that experiencing oral motor interference during encoding

the brand names in the commercials caused a negative affect because the usually running

pronunciation simulations were disturbed, such as any dual task that draws on the same

resources is troubling (Pashler, 1994). Later in the test phase, encountering the advertised

brand names re-activated this feeling of disturbance, while novel names did not, which caused

the present pattern. However, a reversal of advertising effects due to oral interference was not

theoretically predicted and should be interpreted cautiously.

The alternative mechanism that consuming a sugar cube was frustrating or irritating

compared to particularly the popcorn group is unlikely since –if at all– negative mood would

induce a systematic mind-set rendering persuasion effects more unlikely (cf., cognitive tuning,

Schwarz, 2002), as has been shown for brand names (Maheswaran, Mackie, Chaiken, 1992).

However, it was the sugar cube-group that showed advertising effects. Also, participants in

the sugar cube condition could not see what the other group received due to the viewing

partitions in the hall. Also, a possibly greater hunger in this control group is an unlikely

mediator of the present effects, since the target products were not food-related. Finally, the

alternative explanation that eating popcorn compared to eating a sugar cube simply distracts

attention away from the to-be-encoded commercials is ruled out by the fact that chewing gum,

which does not distract vision at all, led to the same inhibition of advertising effects. The

second experiment should map the causal attitudinal undercurrents of these monetary brand

choices in an even more ecologically valid set-up.

Experiment 2

A cinema session was held in a real movie theatre presenting commercials of various

foreign products under oral interference or control conditions. One week later, we assessed


participants’ attitudes towards the brands. In general, attitudes can be assessed by direct

verbal reports, but also via more indirect measures that assess also unconscious attitudinal

determinants of consumer choices that stem from implicit memory (Shapiro & Krishnan,

2001; Strack, Werth, & Deutsch, 2006). For instance, a diffuse familiarity due to implicit

memory that is not necessarily conscious can nevertheless guide preferences and can drive

advertising effects even after long time lags (Shapiro & Krishnan, 2001). One physiologic

measure of implicit memory is electrodermal activity (EDA), that is, the activity of the sweat

glands, such as in lie detecting. Recent research has shown that EDA rises for previously

presented stimuli, even when the participant is not aware of the previous encounter anymore

(de Vries et al., 2010; Morris, Cleary, & Still, 2008; Topolinski, 2012). An important

advantage of such indirect, or implicit, measures is that they are unlikely to be biased by

conscious or strategic processes of the participant, such as conscious recollection of the prior

advertising and resulting reactance against this persuasion attempt (Shapiro & Krishnan,

2001). Moreover, EDA has been shown to be a powerful predictor of sales results (La Barbera

& Tucciarone, 1995). Therefore, we assessed both explicit attitudes via verbal-reports and

implicit memory via EDA in respective sub-samples of this study. We predicted that

participants in the control group would show more positive attitudes towards advertised than

novel options, both for explicit and implicit attitude, while participants with oral motor-

interference would not show such this advertising effect.

Method

Participants. N = 98 (85 female, 13 male; mean age 21, SD = 3) psychology freshmen

in their first week of courses from a university took part for course credit. Two participants

did not attend the test phase, resulting in N = 96.

Materials. Real commercials and images of 36 actually existing foreign products and

their brand names were used, namely Boags, Carling, Dorna, Drench, Fevicol, Finax, Fizz


Diamond, Fritos, Genex, Gyumri, Innis Free, Kinki, Kokanee, Lacvert, Lays, Leonidas,

Lumög, Lurpak, Maeil, Magjia & Shkenca, Margo, Max Havelaar, Montavit, Motilium 10,

Ngan Yin, Nurofen, Palitte, Panapp, Pert Plus, Puhas Loodus, Soyjoy, Stella Artois, Tostitos,

Tsubaki, Veikkaaja, Wacoal. The products were beverages, foods, medicines, perfume,

snacks, toiletries, underwear, and magazines. Pilot studies had shown that these products were

unfamiliar to German participants. The language in the commercial was always non-German.

The commercials were retrieved from various internet resources and were on average 32

seconds in length (ranging from 15 to 82 seconds, SD = 14). The product name was auditorily

mentioned in half of the commercials, counter-balanced across experimental conditions. This

pool was selected from a larger pool of stimuli (Ni = 60) according to moderate valence and

arousal ratings of the commercials in a pilot study (N = 29). The movie in the cinema session

was the animation short film Big Buck Bunny (10 minutes).

Procedure. In the study phase, participants were tested in two groups of around 50

individuals. The procedure of the study session was similar to Experiment 1 except for the

following modifications. The sessions took place in a real movie theatre. Popcorn portions

were 70 gram. This time, the oral interference (popcorn) and control (sugar cube) groups were

separated from each other by seating the participants either in the front or back rows,

respectively (with assignment of condition to front or back counter-balanced across sessions)

and leaving the middle rows empty. Due to the dim light in the cinema hall, neither group

could see what the other received. After instruction and ascertaining that participants had

started consuming the snacks, 6 filler commercials (together 4 minutes in length) were

presented to ensure the sugar cubes had dissolved in the control participants’ mouths. The

longer time of preceding filler commercials (4 minutes) was chosen because participants’ oral

behavior could not readily be observed in the dark room. Therefore, the longer time ensured

that all participants in the control condition had consumed the sugar cube by the time the

target commercials were presented, so that they experienced no oral interference while


watching them. As a pilot study (N = 25) had shown, no participant of an independent sample

was able to hold a sole sugar cube in the mouth without complete dissolving of the cube for

longer than four minutes, even when instructed to keep at least some sugar particles intact as

long as they can. Then, the target commercials were shown (one random half of all products,

counter-balanced across sessions) in a fixed random order (total length 14 minutes) followed

by a short movie. After the movie, participants received a paper-pencil questionnaire with

some filler items with ratings for the movie, mood, and their eating behavior.

For the test phase one week after the study sessions, participants were randomly

assigned to one of the following dependent measures.

Liking ratings. A randomly chosen subsample of n = 56 (28 control, 28 popcorn) was

presented the images of all 36 products in random order on a PC screen for 5 seconds each

and was asked to indicate their spontaneous preference for the product on a 7-point Likert

scale (0 not at all to 6 very much).

Electrodermal activity (EDA). A randomly chosen subsample of n = 40 (20 control,

20 popcorn) was presented the images of all 36 products for 10 seconds each while

electrodermal activity was assessed without any judgmental task. Skin conductance was

measured using two Ag/AgCl surface electrodes placed on the hypothenar eminence of the

palmar surface of the left hand and recorded with a V-Amp 16 amplifier (Brain- Products Inc.,

Richardson, Texas) at 1,000 Hz. Data were stored on an additional PC. As an indicator of

SCR, the difference between the skin conductance level for the 10 s of product image

presentation and a prestimulus baseline of 1 s before stimulus onset was calculated (cf. de

Vries et al., 2010; Morris et al., 2008). This SCR was then collapsed over the 18 advertised

and 18 novel products. In these averaged SCRs, seven participants exhibited electrodermal

activity 2 SDs above/below the sample mean. Their data were discarded (cf., Topolinski,

2012).


Results

Preference ratings. In the control group, advertised products were liked more

(M = 3.03, SE = 0.14) than novel products (M = 2.78, SE = 0.14), t(27) = 2.98, p = .006; d =

0.37. In contrast, the popcorn group did not show this effect (Madvertised = 2.70, SE = 0.13 vs.

Mnovel = 2.71, SE = 0.17, t < 1. However, in a 2 (exposure: advertised, novel products; within)

X 2 (oral motor-interference: yes, no; between) ANOVA the interaction term between

exposure and oral interference was only marginal, F(1, 54) = 2.22, p = .093 (other Fs < 2.7,

n.s.) probably due to the small sample size (Rosenthal, Rosnow, & Rubin, 2000).

EDA. In the sub-sample (n = 40, 7 exclusions, see methods’ section) in which we

assessed EDA we found the following pattern. Participants in the control group showed

higher SCRs to advertised (M = -0.01, SE = 0.1) than to novel products (M = -0.05, SE =

0.01), t(16) = 2.28, p = .037, d = .76, which is the default response to familiarity (Topolinski,

2012). This effect was absent in the popcorn group (Madvertised = -0.04, SE = 0.01 vs. Mnovel = -

0.01, SE = 0.01, t < 1.54, p = .15). This pattern was supported by a 2 (exposure: advertised,

novel products; within) X 2 (oral motor-interference: yes, no; between) ANOVA yielding

only an interaction between exposure and oral interference, F(1, 31) = 7.10, p = .012; ηp2 =

0.19 (other Fs < 1).

Discussion

Both autonomous responses and brand evaluations were increased for advertised

compared to novel products, but only in the control group and not in the group that ate

popcorn while watching the advertising commercials.

Meta-analysis of present results

Although the patterns of findings were consistent across Experiment 1 and 2 and their

different dependent measures, not all individual findings were significant. Hence, we wanted


to test the validity of the overall pattern using two meta-analyses. Since Experiment 1 featured

variables on a categorical, and Experiment 2 variables on an interval scale, we chose the

following two approaches.

Stouffer method. Following Rosenthal‘s (1978) advices we chose the Stouffer

method for combining the present results because we had a small N of studies and t-values

were not available for Experiment 1. The Stouffer methods simply requires to divide the sum

of the standard normal deviates or Zs associated with the one-tailed p-values obtained in each

study by the square root of the number of studies being combined. Because EDA and liking

were assessed on independent samples in Experiment 2 and were not treated as a between-

factor, we treated these two sub-samples as two independent studies. As p-values, we chose

the p-value of the Chi-square test of unequal distribution across oral and control groups in

Experiment 1, and the p-values of the interactions in Experiment 2 (since these interactions

test our crucial prediction). The resulting z = -4.40 was highly significant, p < .00001.

Non-parametric joint analyses. First, we derived a common measure on rank scale

for all studies. For Experiment 1 in which participants had two choices (buying a lotion and

donating for charity), we averaged the likelihood for a given participant of choosing an

advertised product, this average could be -1 (in both choices the participant had chosen a

novel option), 0 (for one of the options the participant had chosen an advertised product), or 1

(in both choices an advertised option was chosen). For Experiment 2, we simply calculated

the difference in liking/EDA for old vs. novel products and assigned -1 to participants for

whom this difference was below zero, and a 1 when this difference was above zero. Thus, for

both experiments the derived measure indicated whether the participant had preferred an

advertised or novel product. Both a Mann-Whitney U-Test, U(277) = 6311.50, p < .00001,

and a Kruskal-Wallis H-Test, H(1) = 27.61, p < 0.0001, found that in control groups the

likelihood was higher to prefer an advertised option than in the oral-interference groups.


General Discussion

In the present experiments, we investigated the role of motor fluency in the mere

exposure effect for advertising (Grimes & Kitchen, 2007). We hypothesized that the driving

mechanism for more positive attitudes for repeated compared to novel brand names is the

efficiency of covert sub-vocalizations, or pronunciation simulations, of brand names, which

run automatically during encountering a name (Stroop, 1935) and are trained when these

names are encountered repeatedly (Topolinski & Strack, 2009b). We prevented the mouth

from covertly simulating name pronunciations and thereby train pronunciation for advertised

brands by letting participants do what is most often done in media situations involving

advertising, consuming snacks.

We found that oral interference indeed obstructed exposure-effects on brand attitudes

(Experiment 2) and even brand choice itself (Experiment 1). While control participants

showed more positive attitudes for and were more likely to spend money on previously

advertised compared to novel products, participants who had eaten popcorn (or chew a gum)

did not show advertising effects. Striking, not only explicit brand attitude, but also

unconscious physiological responses of skin conductance level (La Barbera & Tucciarone,

1995; Shapiro & Krishnan, 2001) were affected.

This pattern of findings show that motor components play a key role in fluency

effects. Converging with other most recent findings, the present studies on advertising brands

show that it is the efficiency of bodily resonances that are automatically and unconsciously

triggered by the brand names that drives preferences (cf., Leder et al., 2013; Sparenberg,

Topolinski, Springer, & Prinz, 2012; Topolinski, Maschmann, Pecher, & Winkielman, 2013).

In the following, we address possible alternative explanations and finally implications for

applied issues.


Alternative explanations

For the present findings, several possible alternative mechanisms could be thought of,

which we will discuss in the following.

Distraction and frustration. As already discussed in Experiment 1 it is possible that

eating popcorn during watching commercials is both more pleasant and more distracting than

the control condition in which participants received a sole sugar cube that was already

dissolved at the time the commercials were presented, which could be the actual cause of

missing advertising effects. However, the activity of chewing a tasteless gum had the same

impact as eating popcorn in Experiment 1, although being less pleasant and not at all

distracting from commercials.

Recognition. It is possible that participants could simply remember the products from

the earlier commercials and chose those products that they recognized (see Janiszewski, 1993;

Shapiro & Krishnan, 2001). The present oral interference simply disturbed this conscious

memory effect. Although such recognition effects were not technically ruled out in the present

experiments, they are unlikely alternative mediators because of the following. First, conscious

recognition of the products is unlikely because participants had also seen several filler

commercials for other products and there was a delay of one week between watching the

commercials and the eventual test phase. Furthermore, generally recollection plays no causal

role in the link between advertising and consumer choice (Janiszewski, 1993; Lodish et al.,

1995). And finally and most importantly, in several experiments dissociating different

memory components, Topolinski (2012) has shown that oral interference impairs both

implicit memory and preference (as in the current studies), but leaves conscious recognition

unaffected.

Brand names or visual appearance of products. Encountering a product does of

course not only entail reading the brand name, but also perceiving the visual appearance of

the product and its packing, for instance color and shape. Thus, maybe these visual cues are


more important in brand choice than simply the name and were affected by the current

manipulations. However, the pattern of the present results strongly suggests that participants

based their choices on the brand name (cf., Fombrun & Shanley, 1990; Maheswaran et al.,

1992). First, in Experiment 1 participants did not see images of the products in the test phase

when chosing lotions and charities, but saw only the brand names. Thus, a visual route was

not even possible. Second, there is no plausible way in which the perceptual encoding of the

visual appearance could be blocked by oral interference: eyes are not affected by a chewing

mouth. Note that Topolinski and Strack (2009b, Experiments 1-2) had found that mere

exposure for images (Chinese ideographs) was not blocked by oral interference. Thus, it was

the brand name and not the products’ visual appearance that drove the current advertising

effects and was blocked by eating popcorn.

Implications for Applied Issues: Snacking and Commercials and the Exclusive

Importance of Brand Names

The present evidence in ecologically valid set-ups, in combination with earlier basic

research on detrimental effects of oral interference on preference and memory for verbal

material (Topolinski, 2012; Topolinski & Strack, 2009b, 2010) suggests that advertising for

novel brands may be futile for snacking cinema audiences, which contradicts common

marketing strategies. Generalizing from the present to other set-ups, it might be speculated

that also commercials in TV (Lodish et al., 1995) would fail to induce a positive attitude

towards advertised novel brands for snacking or even talking TV audience. This possibility

should be addressed in future research.

Given that the current consumer choices so exclusively draw on the brand names

themselves as heuristic cues (Maheswaran et al., 1992) and not on the products’ visual

appearance, and given that the current manipulation targeted particularly on the verbal

system, it might be concluded that in order to avoid such detrimental effects, advertising


should focus more on products’ visual appearance. This might overrun oral interferences on

the brands themselves and might establish implicit memory and thus positive attitudes

towards the product on a visual route. However, the present Experiment 2 did use images of

the whole products, including their design and layout, and still found detrimental effects of

oral interference.Participants in that study could well focus on the visual appearance of the

product, its color, shape, and layout, ignoring the brand name. However, obviously they did

not. This suggests that even when additional perceptual and contextual cues are available,

individuals focus on the brand name, and not on other cues (cf., Fombrun & Shanley, 1990;

Maheswaran et al., 1992).


References

Alter, A. L., & Oppenheimer, D. (2006). Predicting short-term stock fluctuations by using

processing fluency. Proceedings of the National Academy of Sciences, USA, 103, 9369–

9372.

Baker, W. E. (1999). When can affective conditioning and mere exposure directly influence

brand choice. Journal of Advertising, 28(4), 31-46.

Blüher, R., & Pahl. S. (2007). Der 'Mere-Exposure'-Effekt und die Wahl von Produkten: ein

Experiment im Feld [The Mere Exposure Effect and Product Choice: A Field

Experiment]. Zeitschrift für Sozialpsychologie, 38(3), 209-215.

Bornstein, R. F. (1989). Exposure and affect: Overview and meta-analysis of research, 1968–

1987. Psychological Bulletin, 106(2), 265.

de Vries, M., Holland, R. W., Chenier, T., Starr, M. J., & Winkielman, P. (2010). Happiness

cools the warm glow of familiarity: Psychophysiological evidence that mood modulates

the familiarity-affect link. Psychological Science, 21, 321–328.

doi:10.1177/0956797609359878

Elliott, R., & Yannopoulou, N. (2007). The nature of trust in brands: a psychosocial model.

European Journal of Marketing, 41 (9/10), 988 - 998.

Esch, F. R., Möll, T., Schmitt, B., Elger, C. E., Neuhaus, C., & Weber, B. (2012). Brands on

the brain: Do consumers use declarative information or experienced emotions to

evaluate brands? Journal of Consumer Psychology, 22 (1), 75-85.

Fombrun, C., & Shanley, M (1990). What's in a Name? Reputation Building and Corporate

Strategy. The Academy of Management Journal 33(2), 233. doi:10.2307/256324

Grimes, A., & Kitchen, P. (2007). Researching mere exposure effects to advertising -

theoretical foundations and methodological implications. International Journal of

Market Research, 49(2), 191-219.


Janiszewski, C. (1993). Preattentive mere exposure effects. Journal of Consumer Research,

20, 376–393.

Keller, K. L. (2003). Brand Synthesis: The Multidimensionality of Brand Knowledge. Journal

of Consumer Research, 29, 595–600.

La Barbera, P. A., & Tucciarone, J. D. (1995). GSR reconsidered: a behavior approach to

evaluating and improving the sales potency of advertising. Journal of Advertising

Research, 35 (5), 33–53.

Leder, H., Bär, S., & Topolinski, S. (2013). Covert painting simulations influence aesthetic

appreciation of artworks. Psychological Science, 23(12), 1479-1481.

Lodish, L. M., Magid, A., Kalmenson, S., Livelsberger, J., & Lubetkin, B. (1995). How TV

advertising works: a meta-analysis of 289 real world split cable TV advertising

experiments. Journal of Marketing Research, 32 (2), 125-139.

Maheswaran, D., Mackie, D. M., Chaiken, S. (1992). Brand name as a heuristic cue: the

effects of task importance and expectancy confirmation on consumer judgments.

Journal of Consumer Psychology, 1, 317–336.

Matthes, J., Schemer, C., & Wirth, W. (2007). More than meets the eye: Investigating the

hidden impact of brand placements in television magazines. International Journal of

Advertising, 26(4), 477-503.

Moreland, R. L., & Topolinski, S. (2010). The mere exposure phenomenon: A lingering

melody by Robert Zajonc. Emotion Review, 2(4), 329-339.

Morris, A. L., Cleary, A. M., & Still, M. L. (2008). The role of autonomic arousal in feelings

of familiarity. Consciousness and cognition, 17(4), 1378-1385.

Novemsky, N., Dhar, R., Schwarz, N., & Simonson, I. (2007). Preference fluency in

consumer choice. Journal of Marketing Research, 44, 347–356.

Pashler, H. (1994). Dual-task interference in simple tasks: data and theory. Psychological

Bulletin, 116(2), 220.


Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure:

is beauty in the perceiver's processing experience?. Personality and social psychology

review, 8(4), 364-382.

Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective

judgments. Psychological Science, 9(1), 45-48.

Reimann, M., Castaño, R., Zaichkowsky, J., & Bechara, A. (2012). How we relate to brands:

Psychological and neurophysiological insights into consumer–brand relationships.

Journal of Consumer Psychology, 22 (1), 128-142.

Rosenthal, R. (1978). Combining results of independent studies. Psychological Bulletin,

85(1), 185.

Rosenthal, R., Rosnow, R. L., & Rubin, D. B. (2000). Contrasts and effect sizes in behavioral

research: A correlational approach. New York, NY: Cambridge University Press.

Schmitt, B. (2012). The consumer psychology of brands. Journal of Consumer Psychology,

22 (1), 7–17.

Schwarz, N. (2002). Situated cognition and the wisdom of feelings: Cognitive tuning. In L.

Feldman Barrett & P. Salovey (Eds.), The wisdom in feelings (pp.144-166). New York:

Guilford.

Shapiro, S., & Krishnan, H. S. (2001). Memory-based measures for assessing advertising

effects: a comparison of explicit and implicit memory effects. Journal of Advertising,

30 (3), 1–14.

Sparenberg, P., Topolinski, S., Springer, A, & Prinz, W. (2012). Minimal mimicry: Mere

effector matching induces preference. Brain and Cognition, 80 (3), 291-300.

Strack, F., Werth, L., & Deutsch, R. (2006). Reflective and impulsive determinants of

consumer behavior. Journal of Consumer Psychology, 16(3), 205-216.

Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental

Psychology, 12, 242–248.


Topolinski, S. (2010). Moving the eye of the beholder: Motor components in vision determine

aesthetic preference. Psychological Science, 21(9), 1220– 1224.

Topolinski, S. (2012). The sensorimotor contributions to implicit memory, familiarity, and

recollection. Journal of Experimental Psychology: General, 141 (2), 260-281.

Topolinski, S., Likowski, K. U., Weyers, P., & Strack, F. (2009). The face of fluency:

Semantic coherence automatically elicits a specific pattern of facial muscle reactions.

Cognition and Emotion, 23 (2), 260-271.

Topolinski, S., Maschmann, I. T., Pecher, D., & Winkielman, P. (2013). Oral approach-

avoidance: Affective consequences of muscular articulation dynamics. Manuscript

submitted for publication.

Topolinski, S., & Reber, R. (2010). Gaining insight into the „Aha“- experience. Current

Directions in Psychological Science, 19(6), 402-405.

Topolinski, S. & Strack, F. (2009a). The analysis of intuition: Processing fluency and affect in

judgements of semantic coherence. Cognition and Emotion, 23 (8), 1465-1503.

Topolinski, S. & Strack, F. (2009b). Motormouth: Mere exposure depends on stimulus-

specific motor simulations. Journal of Experimental Psychology: Learning, Memory,

and Cognition, 35 (2), 423-33.

Topolinski, S. & Strack, F. (2009c). Scanning the “fringe” of consciousness: What is felt and

what is not felt in intuitions about semantic coherence. Consciousness and Cognition,

18, 608 – 618.

Topolinski, S. & Strack, F. (2010). False fame prevented - avoiding fluency-effects without

judgmental correction. Journal of Personality and Social Psychology, 98(5), 721-733.

Venkatraman, V., Clithero, J. A., Fitzsimons, G. J., & Huettel, S. A. (2012). New scanner data

for brand marketers: How neuroscience can help better understand differences in brand

preferences. Journal of Consumer Psychology, 22 (1), 143-153.


Acknowledgments

This research was supported by the Deutsche Forschungsgemeinschaft (DFG, TO 705/1-1). S.

T. developed the study concept. S. T. and A. L. F. developed the study design of Experiment

1. A. L. F. ran data collection and data analysis of Experiment 1. S. T. and S. L. developed the

study design of Experiment 2. S. L. ran data collection and data analysis of Experiment 2. S.

T. wrote the paper. We thank Amelie Backfisch, Theresa Geiling, Katharina Hass, Elisabeth

Königstein, and Katharina Schatte for their support in data collection. We thank the

Cinemaxx© Würzburg, the Café zum schönen René, and the Standard Würzburg for their

logistic support.


Figure 1: Likelihood of choosing an advertised option in Experiment 1 (error bars are

standard errors).

popcorn in the cinema 1 running head: popcorn and ... · half received 65 gram popcorn or a chewing...

Documents